shunxing1234
commited on
Commit
•
cd0b54c
1
Parent(s):
76504a1
Upload 2 files
Browse files- config.json +42 -0
- configuration.json +1 -0
config.json
ADDED
@@ -0,0 +1,42 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
{
|
2 |
+
"apply_residual_connection_post_layernorm": false,
|
3 |
+
"architectures": [
|
4 |
+
"TeleChatForCausalLM"
|
5 |
+
],
|
6 |
+
"auto_map": {
|
7 |
+
"AutoConfig": "configuration_telechat2.Telechat2Config",
|
8 |
+
"AutoModelForCausalLM": "modeling_telechat2.Telechat2ForCausalLM"
|
9 |
+
},
|
10 |
+
"attention_dropout": 0.0,
|
11 |
+
"attention_softmax_in_fp32": true,
|
12 |
+
"bias_dropout_fusion": true,
|
13 |
+
"bos_token_id": 1,
|
14 |
+
"eos_token_id": 2,
|
15 |
+
"hidden_dropout": 0.0,
|
16 |
+
"hidden_size": 4096,
|
17 |
+
"initializer_range": 0.02,
|
18 |
+
"layer_norm_epsilon": 1e-05,
|
19 |
+
"masked_softmax_fusion": true,
|
20 |
+
"model_type": "telechat",
|
21 |
+
"n_head": 32,
|
22 |
+
"num_key_value_heads":32,
|
23 |
+
"n_inner": null,
|
24 |
+
"n_layer": 30,
|
25 |
+
"offset_alibi": 100,
|
26 |
+
"pad_token_id": 3,
|
27 |
+
"pretraining_tp": 2,
|
28 |
+
"skip_bias_add": true,
|
29 |
+
"skip_bias_add_qkv": false,
|
30 |
+
"slow_but_exact": false,
|
31 |
+
"transformers_version": "4.24.0",
|
32 |
+
"unk_token_id": 0,
|
33 |
+
"use_cache": true,
|
34 |
+
"vocab_size": 131072,
|
35 |
+
"ffn_hidden_size": 12288,
|
36 |
+
"flash_attn":true,
|
37 |
+
"tie_word_embeddings":false,
|
38 |
+
"training_seqlen":8192,
|
39 |
+
"base_seqlen":8192,
|
40 |
+
"seq_length": 32768
|
41 |
+
}
|
42 |
+
|
configuration.json
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
{"framework":"Pytorch","task":"text-generation"}
|